To be suitable for device fabrication, layers of semiconductor materials must have high purity and low density of defects or dislocations. Preparation of quality layers of certain materials, especially compound materials, has been hindered in the past because of a lack of both quality bulk, single crystals, from which individual layers of useful sizes can be sliced, and also of adequate substrates for heteroepitaxial growth of quality layers, from which epitaxial layers can be transferred.
For example, such problems have hindered the development of semiconductors including gallium nitride (GaN); of other Group III nitrides, including, e.g., AlN, InN, GaInN, and other mixed nitrides (referred to herein as “III nitrides”); of certain Group III-V compounds; and of certain other compound materials (e.g., II-VI materials) generally. For many of these materials, semiconductor-quality bulk crystals are not available, and suitable and commercially useful substrates have not been found. A suitable substrate closely matches the crystal properties of the target material to be grown; if these properties do not closely match, the resulting material usually has an unacceptable density of defects and dislocations.
Specifically, in the case of GaN, crystal quality can be improved by pre-treatment of the growth substrates, e.g., by nitridization and other chemical modifications; by growing thin, low temperature buffer layers of other III nitrides, e.g., AlN or GaN, by thermal annealing, and the like. Crystal quality has also been improved by exploiting epitaxial lateral overgrowth (ELO). Despite these advances, available GaN layers still have significant numbers of defects and dislocations.
Such problems have hindered development of alloy semiconductors, e.g., alloys of silicon (Si) and germanium (Ge), Ge layers on Si substrates, and strained Si (sSi) layers. Although crystal quality can be improved by growing these materials on buffer layers having a graded composition, room still remains for improvement.
Clearly, layers and crystals of compound and alloy semiconductors of improved quality are desirable. However, widely applicable processes for doing so are not known in the prior art.